target::TargetBinary< T > Class Template Referenceabstract
Inheritance diagram for target::TargetBinary< T >:
Collaboration diagram for target::TargetBinary< T >:

Public Member Functions

virtual arma::Mat< T > pa ()
 Calculate risk probabilities conditional on exposure. More...
 
virtual arma::Mat< T > p (bool exposure=0)
 
virtual arma::Col< T > loglik (bool indiv=false)
 log likelihood More...
 
virtual arma::Mat< T > score (bool indiv=false)
 score function More...
 
virtual arma::Mat< T > est (arma::Col< T > alpha, const arma::Col< T > &propensity)
 Estimating function for double robust estimator. More...
 
virtual arma::Mat< T > est (arma::Col< T > alpha)
 
void calculate (bool target=true, bool nuisance=true, bool propensity=false) override
 
- Public Member Functions inherited from target::Target< T >
 Target ()
 Constructor.
 
 Target (const arma::Col< T > &y, const arma::Mat< T > &a, const arma::Mat< T > &x1, const arma::Mat< T > &x2, const arma::Mat< T > &x3, const arma::Col< T > &parameter, const arma::Col< T > &weights)
 Target class constructur. More...
 
 Target (const arma::Col< T > &y, const arma::Mat< T > &a, const arma::Mat< T > &x1, const arma::Mat< T > &x2, const arma::Col< T > &parameter, const arma::Col< T > &weights)
 
 Target (const arma::Col< T > &y, const arma::Mat< T > &a, const arma::Mat< T > &x1, const arma::Mat< T > &x2, const arma::Mat< T > &x3, const arma::Col< T > &parameter)
 
 Target (const arma::Col< T > &y, const arma::Mat< T > &a, const arma::Mat< T > &x1, const arma::Mat< T > &x2, const arma::Col< T > &parameter)
 
void weights (const arma::Col< T > &weights)
 
arma::Col< T > weights ()
 
arma::Col< T > A ()
 
arma::Col< T > Y ()
 
arma::Mat< T > X1 ()
 
arma::Mat< T > X2 ()
 
arma::Mat< T > X3 ()
 
void update_data (const arma::Col< T > &y, const arma::Mat< T > &a, const arma::Mat< T > &x1, const arma::Mat< T > &x2, const arma::Mat< T > &x3)
 
void update_par (const arma::Col< T > &parameter)
 update_par - More...
 

Protected Member Functions

virtual arma::Col< T > H ()=0
 
virtual arma::Mat< T > dp ()=0
 

Protected Attributes

arma::Mat< T > pr
 
- Protected Attributes inherited from target::Target< T >
arma::Col< T > nuisance
 
arma::Col< T > target
 
arma::Col< T > propensity
 
arma::Col< T > _response
 
arma::Mat< T > _exposure
 
arma::Mat< T > _x1
 
arma::Mat< T > _x2
 
arma::Mat< T > _x3
 
arma::Col< T > _weights
 

Additional Inherited Members

- Public Attributes inherited from target::Target< T >
arma::Col< T > alpha
 
arma::Col< T > beta
 
arma::Col< T > gamma
 

Detailed Description

template<class T>
class target::TargetBinary< T >

Definition at line 87 of file target.hpp.

Member Function Documentation

◆ est()

template<typename T >
arma::Mat< T > target::TargetBinary< T >::est ( arma::Col< T >  alpha,
const arma::Col< T > &  propensity 
)
virtual

Estimating function for double robust estimator.

\[ U(\alpha; \widehat{\alpha}, \widehat{\beta}, \widehat{\gamma}) = \omega(V)\Big(A-e(V;\widehat{\gamma})\Big)\left (H(\alpha)-p_{0}(V;\widehat{\alpha},\widehat{\beta})\right) \]

  1. Inital estimates \(\widehat{\alpha},\widehat{\beta}\) obtained from MLE.
  2. Similarly, \(\widehat{\gamma}\) obtained from regular asymptotic linear model (e.g., logistic regression MLE).
  3. Plugin estimates to obtain \(\widehat{\omega}_{\mathrm{eff}}\). Also, note that \(p_0\) is calculated wrt initial MLE.
Parameters
alphatarget parameter
propensityexposure propensity weights
Returns
arma::Mat<T>

Definition at line 170 of file target.cpp.

◆ loglik()

template<typename T >
arma::Col< T > target::TargetBinary< T >::loglik ( bool  indiv = false)
virtual

log likelihood

Parameters
indiv- if true return individual log likelihood contributions
Returns
vector of type arma::Vec<T>

Definition at line 142 of file target.cpp.

◆ pa()

template<typename T >
arma::Mat< T > target::TargetBinary< T >::pa ( )
virtual

Calculate risk probabilities conditional on exposure.

\(p_a(V) = E(Y \mid A=a, V), a\in\{0,1\}\)

Returns
arma::mat

Definition at line 125 of file target.cpp.

◆ score()

template<typename T >
arma::Mat< T > target::TargetBinary< T >::score ( bool  indiv = false)
virtual

score function

Parameters
indiv- If true the individual score contributions are returned, otherwise the sum is returned
Returns
arma::Mat

Definition at line 200 of file target.cpp.


The documentation for this class was generated from the following files: